tenzr::numpy¶

Classes¶

Functions¶

Functions Documentation¶

function PromoteTypes¶

TENZR_APIDType PromoteTypes(
    DType a,
    DType b
)

Parameters:

• a The first dtype.
• b The second dtype.
• a Left operand type.
• b Right operand type.

Return:

• The promoted dtype.
• The promoted type.

User-friendly multidimensional array processing layer (tenzr::numpy / tnp). Unlike tenzr::plax, this layer implements standard NumPy ergonomics: automatic type promotion, implicit broadcasting, canonical constructors, and the usual ufuncs, linear algebra, and reductions. Promotes two dtypes to their common NumPy result type. NumPy-style result type of a binary operation.

function BroadcastShapes¶

TENZR_APIShape BroadcastShapes(
    const Shape & a,
    const Shape & b
)

Parameters:

• a The first shape.
• b The second shape.
• a Left operand shape.
• b Right operand shape.

Return:

• The broadcasted shape.
• The broadcast result shape.

Computes the broadcasted shape of two shapes. Broadcast two shapes together.

function BroadcastArrays¶

TENZR_APIBroadcastResult BroadcastArrays(
    const Array & a,
    const Array & b
)

Parameters:

• a The first array.
• b The second array.
• a Left operand.
• b Right operand.

Return:

• The two broadcasted arrays.
• Materialized, shape-matched arrays.

Broadcasts two arrays to a common shape. Broadcast two arrays to a shared contiguous shape.

function PromoteAndBroadcast¶

TENZR_APIBroadcastResult PromoteAndBroadcast(
    const Array & a,
    const Array & b
)

Parameters:

• a The first array.
• b The second array.
• a Left operand.
• b Right operand.

Return:

• The two promoted and broadcasted arrays.
• Materialized, dtype- and shape-matched arrays.

Promotes two arrays to a common dtype and broadcasts them to a common shape. Promote two arrays to a common dtype and broadcast them to a shared shape.

function Zeros¶

TENZR_APIArray Zeros(
    Shape shape,
    DType dtype =DType::kFloat32
)

Parameters:

• shape The output shape.
• dtype The output dtype.
• shape Output shape.
• dtype Element type.

Return:

• The zero-filled array.
• The new array.

Creates an array of zeros with the given shape. Allocate a zero-filled array.

function Ones¶

TENZR_APIArray Ones(
    Shape shape,
    DType dtype =DType::kFloat32
)

Parameters:

• shape The output shape.
• dtype The output dtype.
• shape Output shape.
• dtype Element type.

Return:

• The one-filled array.
• The new array.

Creates an array of ones with the given shape. Allocate a one-filled array.

function Full¶

TENZR_APIArray Full(
    Shape shape,
    double fill_value,
    DType dtype =DType::kFloat32
)

Parameters:

• shape The output shape.
• fill_value The value to fill.
• dtype The output dtype.
• shape Output shape.
• fill_value Value to fill with.
• dtype Element type.

Return:

• The filled array.
• The new array.

Creates an array filled with a constant value. Allocate an array filled with fill_value.

function Arange¶

TENZR_APIArray Arange(
    double start,
    double stop,
    double step =1.0,
    DType dtype =DType::kFloat32
)

Parameters:

• start The start of the interval.
• stop The end of the interval (exclusive).
• step The spacing between values.
• dtype The output dtype.
• start Range start.
• stop Range end (exclusive).
• step Spacing; must be non-zero.
• dtype Element type.

Return:

• The generated array.
• The new array.

Creates evenly spaced values over [start, stop). Evenly spaced values in [start, stop).

function Arange¶

TENZR_APIArray Arange(
    double stop,
    DType dtype =DType::kFloat32
)

Parameters:

• stop The end of the interval (exclusive).
• dtype The output dtype.
• stop Range end (exclusive).
• dtype Element type.

Return:

• The generated array.
• The new array.

Creates evenly spaced values over [0, stop) with a step of 1. Values 0, 1, ..., stop - 1.

function Linspace¶

TENZR_APIArray Linspace(
    double start,
    double stop,
    int64_t num =50,
    DType dtype =DType::kFloat32
)

Parameters:

• start The start of the interval.
• stop The end of the interval (inclusive).
• num The number of samples.
• dtype The output dtype.
• start Range start.
• stop Range end (inclusive).
• num Number of samples; must be positive.
• dtype Element type.

Return:

• The generated array.
• The new array.

Creates num evenly spaced values over the closed interval [start, stop]. num evenly spaced values from start to stop inclusive.

function Eye¶

TENZR_APIArray Eye(
    int64_t N,
    int64_t M =-1,
    int64_t k =0,
    DType dtype =DType::kFloat32
)

Parameters:

• N Number of rows.
• M Number of columns, defaulting to N when negative.
• k Diagonal index (0 is the main diagonal).
• dtype The output dtype.
• N Row count.
• M Column count; defaults to N when negative.
• k Diagonal offset.
• dtype Element type.

Return:

• The identity-like array.
• The new array.

Creates a 2D array with ones on the k-th diagonal and zeros elsewhere. 2-D identity-like matrix with ones on the k-th diagonal.

function Asarray¶

template <typename T >
inline Array Asarray(
    const std::vector< T > & data,
    Shape shape
)

Parameters:

• data The host element values.
• shape The shape of the resulting array.

Template Parameters:

• T The element type.

Return: The array holding data.

Creates an array from a host vector of the given shape.

function array¶

template <typename T >
inline Array array(
    const std::vector< T > & data,
    Shape shape
)

Parameters:

• data The host element values.
• shape The shape of the resulting array.

Template Parameters:

• T The element type.

Return: The array holding data.

Alias of asarray, creating an array from a host vector of the given shape.

function Add¶

TENZR_APIArray Add(
    const Array & a,
    const Array & b
)

Adds two arrays elementwise.

Parameters:

• a Left operand.
• b Right operand.

Return: a + b.

Elementwise addition with promotion and broadcasting.

function Subtract¶

TENZR_APIArray Subtract(
    const Array & a,
    const Array & b
)

Subtracts b from a elementwise.

Parameters:

• a Left operand.
• b Right operand.

Return: a - b.

Elementwise subtraction with promotion and broadcasting.

function Multiply¶

TENZR_APIArray Multiply(
    const Array & a,
    const Array & b
)

Multiplies two arrays elementwise.

Parameters:

• a Left operand.
• b Right operand.

Return: a * b.

Elementwise multiplication with promotion and broadcasting.

function Divide¶

TENZR_APIArray Divide(
    const Array & a,
    const Array & b
)

Divides a by b elementwise.

Parameters:

• a Dividend.
• b Divisor.

Return: a / b.

Elementwise division with promotion and broadcasting.

function Power¶

TENZR_APIArray Power(
    const Array & a,
    const Array & b
)

Raises a to the power b elementwise.

Parameters:

• a Base.
• b Exponent.

Return: a ** b.

Elementwise exponentiation with promotion and broadcasting.

function Maximum¶

TENZR_APIArray Maximum(
    const Array & a,
    const Array & b
)

Takes the elementwise maximum of two arrays.

Parameters:

• a Left operand.
• b Right operand.

Return: Elementwise maximum.

Elementwise maximum with promotion and broadcasting.

function Minimum¶

TENZR_APIArray Minimum(
    const Array & a,
    const Array & b
)

Takes the elementwise minimum of two arrays.

Parameters:

• a Left operand.
• b Right operand.

Return: Elementwise minimum.

Elementwise minimum with promotion and broadcasting.

function Remainder¶

TENZR_APIArray Remainder(
    const Array & a,
    const Array & b
)

Computes the elementwise remainder of a divided by b.

Parameters:

• a Dividend.
• b Divisor.

Return: a % b.

Elementwise remainder with promotion and broadcasting.

function Atan2¶

TENZR_APIArray Atan2(
    const Array & a,
    const Array & b
)

Computes the elementwise two-argument arctangent.

Parameters:

• a First operand.
• b Second operand.

Return: atan2(a, b).

Elementwise two-argument arctangent with promotion and broadcasting.

function Equal¶

TENZR_APIArray Equal(
    const Array & a,
    const Array & b
)

Compares two arrays for elementwise equality.

Parameters:

• a Left operand.
• b Right operand.

Return: Boolean array.

Elementwise equality comparison.

function NotEqual¶

TENZR_APIArray NotEqual(
    const Array & a,
    const Array & b
)

Compares two arrays for elementwise inequality.

Parameters:

• a Left operand.
• b Right operand.

Return: Boolean array.

Elementwise inequality comparison.

function Less¶

TENZR_APIArray Less(
    const Array & a,
    const Array & b
)

Tests whether a is elementwise less than b.

Parameters:

• a Left operand.
• b Right operand.

Return: Boolean array.

Elementwise less-than comparison.

function LessEqual¶

TENZR_APIArray LessEqual(
    const Array & a,
    const Array & b
)

Tests whether a is elementwise less than or equal to b.

Parameters:

• a Left operand.
• b Right operand.

Return: Boolean array.

Elementwise less-than-or-equal comparison.

function Greater¶

TENZR_APIArray Greater(
    const Array & a,
    const Array & b
)

Tests whether a is elementwise greater than b.

Parameters:

• a Left operand.
• b Right operand.

Return: Boolean array.

Elementwise greater-than comparison.

function GreaterEqual¶

TENZR_APIArray GreaterEqual(
    const Array & a,
    const Array & b
)

Tests whether a is elementwise greater than or equal to b.

Parameters:

• a Left operand.
• b Right operand.

Return: Boolean array.

Elementwise greater-than-or-equal comparison.

function Sin¶

TENZR_APIArray Sin(
    const Array & a
)

Computes the elementwise sine.

Parameters:

• a Input array.

Return: sin(a).

Elementwise sine.

function Cos¶

TENZR_APIArray Cos(
    const Array & a
)

Computes the elementwise cosine.

Parameters:

• a Input array.

Return: cos(a).

Elementwise cosine.

function Tan¶

TENZR_APIArray Tan(
    const Array & a
)

Computes the elementwise tangent.

Parameters:

• a Input array.

Return: tan(a).

Elementwise tangent.

function Exp¶

TENZR_APIArray Exp(
    const Array & a
)

Computes the elementwise exponential.

Parameters:

• a Input array.

Return: exp(a).

Elementwise exponential.

function Log¶

TENZR_APIArray Log(
    const Array & a
)

Computes the elementwise natural logarithm.

Parameters:

• a Input array.

Return: log(a).

Elementwise natural logarithm.

function Sqrt¶

TENZR_APIArray Sqrt(
    const Array & a
)

Computes the elementwise square root.

Parameters:

• a Input array.

Return: sqrt(a).

Elementwise square root.

function Rsqrt¶

TENZR_APIArray Rsqrt(
    const Array & a
)

Computes the elementwise reciprocal square root.

Parameters:

• a Input array.

Return: 1 / sqrt(a).

Elementwise reciprocal square root.

function Abs¶

TENZR_APIArray Abs(
    const Array & a
)

Computes the elementwise absolute value.

Parameters:

• a Input array.

Return: |a|.

Elementwise absolute value.

function tanh¶

TENZR_APIArray tanh(
    const Array & a
)

Computes the elementwise hyperbolic tangent.

Parameters:

• a Input array.

Return: tanh(a).

Elementwise hyperbolic tangent.

function Ceil¶

TENZR_APIArray Ceil(
    const Array & a
)

Rounds each element up to the nearest integer.

Parameters:

• a Input array.

Return: ceil(a).

Elementwise ceiling.

function Floor¶

TENZR_APIArray Floor(
    const Array & a
)

Rounds each element down to the nearest integer.

Parameters:

• a Input array.

Return: floor(a).

Elementwise floor.

function Round¶

TENZR_APIArray Round(
    const Array & a
)

Rounds each element to the nearest integer.

Parameters:

• a Input array.

Return: round(a).

Elementwise round-to-nearest.

function Sign¶

TENZR_APIArray Sign(
    const Array & a
)

Returns the elementwise sign (-1, 0, or 1).

Parameters:

• a Input array.

Return: sign(a).

Elementwise sign.

function Negative¶

TENZR_APIArray Negative(
    const Array & a
)

Negates each element.

Parameters:

• a Input array.

Return: -a.

Elementwise negation.

function Clip¶

TENZR_APIArray Clip(
    const Array & a,
    const Array & a_min,
    const Array & a_max
)

Clamps each element to the closed interval [a_min, a_max].

Parameters:

• a Input array.
• a_min Lower bound.
• a_max Upper bound.

Return: The clamped array.

Clamp elements into [a_min, a_max] with promotion and broadcasting across all three operands.

function Matmul¶

TENZR_APIArray Matmul(
    const Array & a,
    const Array & b
)

Matrix-multiplies two arrays, broadcasting batch dimensions.

Parameters:

• a Left operand.
• b Right operand.

Return: a @ b.

Matrix multiplication.

function Dot¶

TENZR_APIArray Dot(
    const Array & a,
    const Array & b
)

Computes the dot product of two arrays.

Parameters:

• a Left operand.
• b Right operand.

Return: The dot product or matmul result.

Vector dot product for rank-1 operands, otherwise matrix multiplication.

function Transpose¶

TENZR_APIArray Transpose(
    const Array & a,
    const std::vector< int64_t > & axes ={}
)

Permutes the array axes, reversing them when axes is empty.

Parameters:

• a Input array.
• axes Optional target axis order; reverses when empty.

Return: The transposed view.

Reverse the axes, or apply an explicit permutation.

function Reshape¶

TENZR_APIArray Reshape(
    const Array & a,
    const Shape & new_shape
)

Returns an array with the requested shape.

Parameters:

• a Input array.
• new_shape Target shape.

Return: The reshaped array.

Reshape without changing the element count.

function Squeeze¶

TENZR_APIArray Squeeze(
    const Array & a,
    std::optional< int64_t > axis =std::nullopt
)

Removes size-1 dimensions, either at axis or all of them.

Parameters:

• a Input array.
• axis Optional axis to squeeze; all unit dimensions when empty.

Return: The squeezed view.

Remove unit dimensions.

function ExpandDims¶

TENZR_APIArray ExpandDims(
    const Array & a,
    int64_t axis
)

Inserts a new size-1 dimension at axis.

Parameters:

• a Input array.
• axis Insertion position.

Return: The expanded view.

Insert a unit dimension at axis.

function Concatenate¶

TENZR_APIArray Concatenate(
    const std::vector< Array > & arrays,
    int64_t axis =0
)

Concatenates arrays along an existing axis.

Parameters:

• arrays Arrays to join, all with the same rank and matching non-axis dimensions.
• axis Concatenation axis.

Return: The concatenated array.

Join arrays along an existing axis, promoting to a common dtype.

function Stack¶

TENZR_APIArray Stack(
    const std::vector< Array > & arrays,
    int64_t axis =0
)

Stacks arrays along a new axis.

Parameters:

• arrays Arrays to stack, all with the same shape.
• axis Position of the new axis.

Return: The stacked array.

Join arrays along a new axis.

function Sum¶

TENZR_APIArray Sum(
    const Array & a,
    std::optional< std::vector< int64_t > > axis =std::nullopt,
    bool keepdims =false
)

Sums elements, optionally over the given axes.

Parameters:

• a Input array.
• axis Axes to reduce; all when empty.
• keepdims Whether reduced axes are retained.

Return: The reduced array.

Sum over the given axes.

function Mean¶

TENZR_APIArray Mean(
    const Array & a,
    std::optional< std::vector< int64_t > > axis =std::nullopt,
    bool keepdims =false
)

Computes the arithmetic mean, optionally over the given axes.

Parameters:

• a Input array.
• axis Axes to reduce; all when empty.
• keepdims Whether reduced axes are retained.

Return: The reduced array.

Arithmetic mean over the given axes.

function Prod¶

TENZR_APIArray Prod(
    const Array & a,
    std::optional< std::vector< int64_t > > axis =std::nullopt,
    bool keepdims =false
)

Multiplies elements, optionally over the given axes.

Parameters:

• a Input array.
• axis Axes to reduce; all when empty.
• keepdims Whether reduced axes are retained.

Return: The reduced array.

Product over the given axes.

function Max¶

TENZR_APIArray Max(
    const Array & a,
    std::optional< std::vector< int64_t > > axis =std::nullopt,
    bool keepdims =false
)

Computes the maximum, optionally over the given axes.

Parameters:

• a Input array.
• axis Axes to reduce; all when empty.
• keepdims Whether reduced axes are retained.

Return: The reduced array.

Maximum over the given axes.

function Min¶

TENZR_APIArray Min(
    const Array & a,
    std::optional< std::vector< int64_t > > axis =std::nullopt,
    bool keepdims =false
)

Computes the minimum, optionally over the given axes.

Parameters:

• a Input array.
• axis Axes to reduce; all when empty.
• keepdims Whether reduced axes are retained.

Return: The reduced array.

Minimum over the given axes.

function Var¶

TENZR_APIArray Var(
    const Array & a,
    std::optional< std::vector< int64_t > > axis =std::nullopt,
    bool keepdims =false
)

Computes the variance, optionally over the given axes.

Parameters:

• a Input array.
• axis Axes to reduce; all when empty.
• keepdims Whether reduced axes are retained.

Return: The reduced array.

Variance over the given axes.

function StdDev¶

TENZR_APIArray StdDev(
    const Array & a,
    std::optional< std::vector< int64_t > > axis =std::nullopt,
    bool keepdims =false
)

Computes the standard deviation, optionally over the given axes.

Parameters:

• a Input array.
• axis Axes to reduce; all when empty.
• keepdims Whether reduced axes are retained.

Return: The reduced array.

Standard deviation over the given axes.

function Where¶

TENZR_APIArray Where(
    const Array & condition,
    const Array & x,
    const Array & y
)

Selects elements from x or y according to condition.

Parameters:

• condition Boolean selector.
• x Value where true.
• y Value where false.

Return: The selected array.

Select x where condition is true and y otherwise.

function Take¶

TENZR_APIArray Take(
    const Array & a,
    const Array & indices,
    int64_t axis =0
)

Takes elements from a along axis using the given indices.

Parameters:

• a Source array.
• indices Index array.
• axis Axis to gather along.

Return: The gathered array.

Gather entries of a along axis at the given indices.